CN106030031B

CN106030031B - Control shaft bottom sub-assembly follows the computer implemented method and system in planning pit shaft path

Info

Publication number: CN106030031B
Application number: CN201380081029.0A
Authority: CN
Inventors: J·D·戴克斯特拉; 孙之杰
Original assignee: Halliburton Energy Services Inc
Current assignee: Halliburton Energy Services Inc
Priority date: 2013-12-06
Filing date: 2013-12-06
Publication date: 2019-11-19
Anticipated expiration: 2033-12-06
Also published as: US20160265334A1; AU2017208259A1; AU2013406720A1; AR098646A1; MX2016006825A; CN106030031A; NO347772B1; CA2930384C; GB2534793A; NO20160832A1; AU2017208259B2; RU2641054C2; GB2534793B; GB201608206D0; US10794168B2; CA2930384A1; WO2015084401A1; RU2016118521A

Abstract

The technology that planning pit shaft path is followed for controlling shaft bottom sub-assembly (BHA) comprises determining that the sensor measurement from the BHA；BHA dynamic model is determined based on the sensor measurement from the BHA；Determine the weighted factor for corresponding to Target For Drilling；Determination includes the objective function of the Target For Drilling weighted by the weighted factor and one or more constraints；Determine that the control to the BHA for meeting the objective function and one or more of constraints inputs；The BHA is applied to by control input.

Description

Control shaft bottom sub-assembly follows the computer implemented method and system in planning pit shaft path

Technical field

This disclosure relates to the automatic management of the wellbore operations from subterranean layer production oil gas.

Background technique

(such as oily gentle) probing of oil gas is usually directed to and can set to the probing for the subterranean depth for assigning underground several thousand feet Standby operation.These long-range distances of downhole drilling apparatus combine uncertain downhole operations condition and vibration probing disturbance to exist Many challenges are formed when the track of accurate control pit shaft.The usually adjacent wellbore for complicating these problems (is close to that sometimes This) presence, limitation probing error tolerance.Drilling operation is usually from the well being located at or near shaft bottom sub-assembly (BHA) Lower sensor collects measurement to detect the various conditions of related probing, such as position of pit shaft track and angle, the characteristic of rock stratum, Pressure, temperature, sound, radiation etc..These sensor measurement datas are usually transferred to ground, and human operator is in Ground analysis institute Data are stated to adjust downhole drilling apparatus.But sensor measurement may inaccuracy, delay or infrequently, limitation use these The validity of measurement.In general, human operator has to control drilling operation using the estimation of the best-guess of pit shaft track.

Summary of the invention

According to the first aspect of the invention, it is real to provide a kind of computer that control shaft bottom sub-assembly follows planning pit shaft path Applying method, which comprises

Determine the sensor measurement from the shaft bottom sub-assembly；

Shaft bottom sub-assembly dynamic model is determined based on the sensor measurement from the shaft bottom sub-assembly；

Determine the weighted factor for corresponding to Target For Drilling；

Determination includes the objective function of the Target For Drilling weighted by the weighted factor and one or more constraints；

Determine that the control to the shaft bottom sub-assembly for meeting the objective function and one or more of constraints inputs； With

Control input is applied to the shaft bottom sub-assembly；

Wherein determine the weighted factor for corresponding to Target For Drilling further include:

In the sensor measurement based on the shaft bottom sub-assembly dynamic model or from the shaft bottom sub-assembly extremely A few determining weighted factor.

According to the second aspect of the invention, a kind of system that control shaft bottom sub-assembly follows planning pit shaft path is provided, Include:

First assembly, rest on the ground or Near Ground；

Shaft bottom sub-assembly is at least partly placed in the pit shaft at subterranean zone or near subterranean zone, the shaft bottom Sub-assembly is associated at least one sensor；With

Controller, is communicatively coupled to the first assembly and the shaft bottom sub-assembly, the controller can operate with It carries out including following operation:

Determine the sensor measurement from the shaft bottom sub-assembly；

Determine the weighted factor for corresponding to Target For Drilling；

Control input is applied to the shaft bottom sub-assembly；

According to the third aspect of the invention we, it provides a kind of to include the non-temporary of at least one computer program code of instruction When property computer readable storage medium, described instruction operate that at least one processor is promoted to carry out for controlling when executed Shaft bottom sub-assembly follows the operation in planning pit shaft path, and the operation includes:

Determine the sensor measurement from the shaft bottom sub-assembly；

Determine the weighted factor for corresponding to Target For Drilling；

Control input is applied to the shaft bottom sub-assembly；

Detailed description of the invention

The example that Fig. 1 illustrates at least part of implementation of the wellbore system under downhole operations background；

The example that Fig. 2 illustrates the process flow of the PREDICTIVE CONTROL based on model, in response to the change condition in pit shaft Dynamic adjustment weighted factor；

Fig. 3 illustrates the probabilistic 3 dimension example of the correlation between the different directions in pit shaft track；

Fig. 4 A and 4B diagram determine the example n in the anticollision direction for weighted factor adjustment；

Fig. 5 is the flow chart of the example of the program of weight adjustment and synthesis；

Fig. 6 is the flow chart for the example procedure for carrying out the PREDICTIVE CONTROL based on model of BHA；

Fig. 7 is based on BHA dynamic model or at least one of the sensor measurement from BHA determines at least one weighting The flow chart of the example of the further details of the factor；

Fig. 8 is at least one determining weighted factor and determines the objective function weighted by least one described weighted factor The flow chart of the example of further details；

Fig. 9 is determining objective function and determines the further details that the control for the BHA for meeting the objective function inputs The flow chart of example；And

Figure 10 is the block diagram of the example for the control system that some examples can operate on it.

Specific embodiment

The disclosure substantially describes to carry out pit shaft drilling operation by making the PREDICTIVE CONTROL decision based on model for BHA Automation control.Particularly, it describes based on the change condition dynamic adjustment BHA control input in pit shaft when different Between emphasize the technologies of different Target For Drillings.Any adequate information source can be used to determine, such as in change condition in pit shaft Sensor measurement, the prediction based on model and/or pit shaft planning information.

It may be in response in pit shaft altered condition adjustment BHA control input and emphasize that (or not emphasizing) and probing are grasped with selectivity Make associated one or more targets.As example, the target may relate to reduce and plan the deviation in pit shaft path, subtract The input energy consumption of small BHA, or any other appropriate target in relation to drilling operation.During drilling operation, change in pit shaft Condition (for example, different rock-layers, the different shape part for planning pit shaft path etc.) can lead to different target to be become in different time It is more important or less important to maintain whole efficiency and cost-effectiveness drilling operation.

In some instances, one or more targets can be combined in single overall goals function, wherein using one or Multiple weighted factors emphasize different target by not same amount.It can implement the adaptability of BHA control input by adjustment weighted factor Matter emphasizes the different target in objective function with selectivity, and solves the BHA control input for meeting overall goals function.Weighting The factor can automatically adapt to the variation of the condition in pit shaft.As example, weighted factor can automatically adapt in pit shaft track not There is the adjacent well bore or well for constituting collision threat in the variation of certainty amount, the different angle along planning pit shaft path and turning In cylinder or surrounding may other conditions relevant to directionality drilling system.

In generalization is implemented, a kind of control shaft bottom sub-assembly (BHA) follows the computer-implemented side in planning pit shaft path Method, which comprises determine the sensor measurement from the BHA；It is determined based on the sensor measurement from the BHA BHA dynamic model；Determine the weighted factor for corresponding to Target For Drilling；Determination includes the Target For Drilling weighted by the weighted factor With the objective function of one or more constraint；Determine the control for meeting the BHA of the objective function and one or more of constraints System input；Control input is applied to the BHA.

Other generalization are implemented including corresponding computer system, equipment and are recorded in one or more Computer Memory Units On computer program, be respectively configured to carry out the movement of method.One or more system for computer can be configured to carry out Operation is to be acted.One or more computer programs, which can be configured to rely on, to be included instruction and carries out specific operation or dynamic Make, described instruction promotes equipment to be acted when being executed by data processing equipment.

In it can combine the first aspect that any one generalization is implemented, determine that the weighted factor for corresponding to Target For Drilling also wraps Include: based on the BHA dynamic model or at least one of the sensor measurement from the BHA determines weighted factor.

In the second aspect in terms of can combining any one previously, the sensor based on BHA dynamic model or from BHA is surveyed Amount at least one of determine weighted factor comprise determining that measurement pit shaft track uncertain, the described pit shaft shape or At least one of collision elimination information；

At least one of shape or collision elimination information of uncertain, the described pit shaft of pit shaft track based on measurement Determine weight；Weighted factor is synthesized with by the weight.

In the third aspect in terms of can combining any one previously, determine that the uncertainty of the pit shaft track of measurement includes true Covariance value between the multiple orientation values and tilting value of the fixed pit shaft track.

In the fourth aspect in terms of can combining any one previously, determine that the weighted factor for corresponding to Target For Drilling is included in The uncertain of the pit shaft track of measurement is wherein being reinforced to the control of the BHA on increased direction from previous measurement time Constraint in system input.

In the 5th aspect in terms of can combining any one previously, the constraint reinforced in the control input of the BHA includes The value added of determination weighted factor associated with the control input of the BHA.

Can combine any one previously aspect the 6th aspect in, the Target For Drilling include with plan pit shaft path it is pre- Deviation is surveyed, and determines that the weighted factor for corresponding to Target For Drilling includes surveying in the uncertain of the pit shaft track of measurement from previous The amount time wherein loosen on increased direction and the prediction deviation in the planning pit shaft path on constraint.

In the 7th aspect in terms of can combining any one previously, loosen and the pact on the prediction deviation in planning pit shaft path Beam includes the decreasing value of weighted factor determining and associated with the planning prediction deviation in pit shaft path.

In the eighth aspect in terms of can combining any one previously, determine that the shape of the pit shaft includes determining planning pit shaft The radius of curvature of the further part in path.

Can combine any one previously aspect the 9th aspect in, the Target For Drilling include with plan pit shaft path it is pre- Deviation is surveyed, and determines that weighted factor includes the radius of curvature in the further part in the planning pit shaft path from previous measurement time Reduce on the wherein direction of reduction and the constraint on the prediction deviation in the planning pit shaft path.

In the tenth aspect in terms of can combining any one previously, determine that collision elimination information includes determination and another pit shaft Collision most probable in the direction wherein occurred.

In the tenth one side in terms of can combining any one previously, the Target For Drilling includes and planning pit shaft path Prediction deviation, and determine that weighted factor includes reinforcing on the direction that the collision most probable with another pit shaft occurs wherein With the constraint on the prediction deviation in the planning pit shaft path.

Any one can combined previously in the 12nd aspect of aspect, multiple orientation values of the pit shaft track are being determined and incline Covariance value between inclined value further include: determine the received multiple azimuthal measuremenies of sensor from the BHA and inclination measurement；With Determine the received multiple azimuthal measuremenies of sensor from the BHA and the covariance value between inclination measurement.

Any one can combined previously in the 13rd aspect of aspect, multiple orientation values of the pit shaft track are being determined and incline Covariance value between inclined value further include: multiple bearing predictions and tilt prediction are determined based on the BHA dynamic model；Be based on The BHA dynamic model determines the covariance value between multiple bearing predictions and tilt prediction.

In the fourteenth aspect in terms of can combining any one previously, determine that objective function comprises determining that and the planning The prediction future deviation in pit shaft path；Determine the prediction future cost that control input is applied to BHA；Added with determining by described Weight factor weighting, and the prediction for predicting the following deviation and the control input is applied to BHA in the planning pit shaft path The weighted array of future cost.

In the 15th aspect in terms of can combining any one previously, determine that weighted factor comprises determining that and the planning First weighted factor of the following deviation of the prediction in pit shaft path；Control is inputted to the prediction future cost for being applied to BHA with determining The second weighted factor.

In the 16th aspect in terms of can combining any one previously, the control to BHA for meeting the objective function is determined System input include determine on the follow-up time period minimize with it is described planning pit shaft path prediction future deviation and will described in The control to BHA that control input is applied to the weighted array of the prediction future cost of BHA inputs.

Any one can combined previously in the 17th aspect of aspect, will control input be applied to prediction future of BHA at This includes the prediction energy consumption of the BHA.

In the 18th aspect in terms of can combining any one previously, further includes: determine that the candidate control to BHA inputs； Prediction pit shaft track is determined with BHA dynamic model based on the candidate control input to BHA；With based on prediction pit shaft track and planning The prediction future deviation of deviation determination and the planning pit shaft path between pit shaft path.

In the 19th aspect in terms of can combining any one previously, determine that the control input of BHA includes determining that first is curved At least one of angle control, the control of the second bent angle, the control of the first packer or the control of the second packer.

In the 20th aspect in terms of can combining any one previously, further includes: determine the biography of the update from the BHA Sensor measurement；The BHA dynamic model updated is determined based on the sensor measurement of the update from the BHA；Based on the update BHA dynamic model or at least one of the sensor measurement of update from the BHA determine the weighted factor updated and The objective function of update；With the weighted factor based on the update automatically adjust meet update objective function to BHA's Control input.

In the 20th one side in terms of can combining any one previously, determine between multiple azimuthal measuremenies and inclination measurement Covariance value further include determine the uncertainty value in two different directions from the pit shaft track between it is mutual It closes.

Each implementation according to the disclosure for the control system of shaft bottom probing may not include including one or it is some with Lower feature.For example, the system can improve the stability and robustness of drilling operation.Particularly, presently disclosed technology It can support the more accurate of pit shaft track and accurately control, but regardless of the variation and unpredictable situation in wellbore environment.

For example, if the specific part of pit shaft generates bigger error and more uncertain measurements, it would be possible that expectation will more Multiple weighing value is placed in the target for reducing input energy, therefore constrains input may not accurately reflect pit shaft in the track of measurement The real trace time during use more conservative probing.As another example, if the specific part tool in planning pit shaft path There is zig zag, it would be possible that less weight is placed on the target for remaining close to planning pit shaft path during these times by expectation On, to allow to have more room in zig zag and avoid limitation input (for example, excessively high if remaining close to planning path It is expensive or even infeasible).Different target is more or less emphasized by the different time adaptability during drilling operation, this The text technology can be supported more effectively and more accurate drilling operation, but regardless of the change condition in pit shaft.

In some instances, BHA dynamic model can be used to generate the prediction of the following pit shaft track, and can be based on prediction Actively adjust BHA control input in pit shaft track.BHA dynamic can be updated when newly being measured and when receiving new control input Model, to support less prediction error of pit shaft track.These can be used to predict and plan pit shaft routing information for the system And/or other information expects the change in future of pit shaft and actively adjusts drilling operation.

For example, the system can be changed based on the expectation in pit shaft and be increased or decreased selected by one or more targets Weighted factor, and automatically determine the BHA control input and one or more constraints for meeting the weighted target of adjustment.Meet institute Stating target may include that or can wrap such as optimizing (for example, minimizing cost function, maximally utilizing rate function) Include find close to optimal solution suboptimal solution (for example, it is contemplated that numerical radius etc. of computation complexity), or may include meet with The relevant other appropriate targets of drilling program.Meet the target BHA control input determination can be related to meet one or Multiple constraints, such as maximum bent angle, maximum available power etc..

Model and constraint determine which optimum control inputs and associated future BHA is moved based on by minimizing objective function State and the form for determining all feasible future BHA behaviors.

The subsurface environment around BHA in pit shaft is usually complicated system.In some instances, the system may wrap Include at least four control variable and 12 measurements.Conventional control strategy can not be easily for a variety of causes for including following reason It is applied to BHA system.Interactions between difference input and output may be strong and uncertain, such as inclination measurement It may depend on most of control variables, such as two bent angles and packer inflation.In such scene, expected performance is being realized When conventional design techniques such as proportional integral differential (PID) may be limited.Such as, if it is desired to objective function has optimal Solution, then PID controller can cannot achieve desired optimal performance.Another difficulty is that output quantity is likely larger than input quantity, And it can't always be the interaction for being clear how to unlock between specific input and output.This, which may cause, makes BHA input control Design the complicated and numerous option complicated.In such scene, the performance of drilling operation generally depends on control system The tuning technical ability of designer (it can suffer from personal error).Another difficulty that the amount of measurement is greater than the amount of control variable is In many cases, all measurements are likely difficult to track its object of planning value without meeting with some deviations.Such deviation may make At the uncertainty for how controlling pit shaft track, so as to cause the excessively aggressive control for the different outputs for causing to compete with each other System.For example, this may cause in remote inclination sensor has if nearby inclination sensor needs larger bent angle More multiple error and uncertainty.In some scenes, this may cause the stabilization tolerance of reduction, so that drilling operation is more difficult to essence Really control.

Technology described herein provides a kind of control strategy based on the PREDICTIVE CONTROL (MPC) based on model, supports to adjust Complicated BHA system, in some instances it may even be possible to the system with strong interaction, while meeting the entirety of (such as optimization) drilling operation Objective function and any associated constraints.In addition, ambient enviroment and design specification are fast during directional drilling operation wherein In the scene of speed variation, adaptability weight tuning algorithm is carried out in combination with MPC strategy to realize more steady and accurately control.

The details that one or more is implemented is illustrated in attached drawing and in being described below.Other spies will be understood from description and figure Sign, objects and advantages.

Fig. 1 illustrates a part of an implementation of the slanted well bore system 100 according to the disclosure.Although being shown as pitch system (for example, there is orientation, horizontal or arc pit shaft), but system can only include Relative vertical pit shaft (e.g., including normally drill Change) and other types of pit shaft (for example, branch well cylinder, well pattern pit shaft and other pit shafts).In addition, although illustrated as on ground On face, but system 100 can be located in seabed or water environment.In general, slanted well bore system 100 enters one or more subterranean layers And provide the simpler and more efficient production for the oil gas being located in these subterranean layers.In addition, slanted well bore system 100 allows Simpler and more efficient pressure break or simulated operation.As illustrated in fig. 1, slanted well bore system 100 includes being deployed in ground 102 On probing sub-assembly 104.Probing sub-assembly 104 can be used to form from ground 102 and pass through one or more geology in ground The vertical bore part 108 that layer extends.One or more subterranean layers (such as producing formation 126) are located at 102 lower section of ground.As incited somebody to action Illustrate in more detail below, one or more wellbore casings (such as surface casing 112 and intermediate sleeve 114) are mountable vertical In at least part of pit shaft part 108.

In some implementations, probing sub-assembly 104 can be deployed on water body rather than ground 102.For example, in some implementations In, ground 102 can be the lower ocean that can find hydrocarbon-bearing formation, bay, ocean or any other water body.In brief, it mentions And ground 102 includes both land and water surface and imagines from the formation of either one or two position and/or generation one or more A slanted well bore system 100.

In general, probing sub-assembly 104 can be any appropriately combined part or drilling machine for forming pit shaft in the ground.Probing group Traditional technology can be used to form these pit shafts (such as vertical bore part 108) or can be used non-traditional or new in component 104 Clever technology.In some implementations, rotary drilling equipment can be used to form these pit shafts in probing sub-assembly 104.Rotary drilling is set It is standby to be known and be made of drill string 106 and shaft bottom sub-assembly (BHA) 118.In some implementations, sub-assembly 104 is drilled It can be made of rotary rig.Slewing on such a rotary rig can be by for forming the component of bit, institute Stating drill bit then forms pit shaft (such as vertical bore part 108) deeper and deeperly to underground.Slewing is by many components (being not entirely shown herein) composition promotes to turn power from motive power to be moved to drill bit itself.Motive power supplies power to turn Disk or top Direct Driving System then supply rotary power to drill string 106.Drill string 106 is usually in shaft bottom sub-assembly 118 It is attached to drill bit.It is attached to the majority (if not all) of change carrying 106 weight of drill string of lifting appliance, but allows drill string 106 rotate freely.

Drill string 106 is usually made of heavy-duty steel pipeline section, is screw thread, so that they can lock together.Under drilling pipe Side is one or more drill collars, heavier than drilling pipe, thicker and harder.The help of screw thread drill collar increases weight above drill bit Drill bit is allowed to drill one or more geological stratifications to ensure that there are enough downward pressures on drill bit to drill string 106.It is any The quantity and property of drill collar on specific rotary rig can be according to the conditions down-hole changes undergone while probing.

Drill bit is usually located in shaft bottom sub-assembly 118 or is attached to shaft bottom sub-assembly 118, the shaft bottom sub-assembly 118 In the downhole end of drill string 106.Drill bit, which is mainly responsible for, to be contacted and drills with the material (for example, rock) in one or more geological stratifications This material.According to the disclosure, bite type can be according to the type selection of the geological stratification met in probing.For example, the probing phase Between the different geological stratifications that suffer from may need to realize maximum drilling efficiency using different drill bits.Drill bit may be because in stratum These differences change because of the abrasion of drill bit experience.It is big although these details are for the disclosure and non-key There are the drill bits of four seed types for cause, are respectively suitable for specified conditions.The drill bit of four kinds of most common types include: delay or drag bit, Steel rotary drilling-head, polycrystalline diamond compact bit and diamond bit.Regardless of selected specific bit, " landwaste " it is continuous It is crucial for removing for rotary drilling.

The circulatory system of rotary drilling operation (such as probing sub-assembly 104) can be the additional assemblies of probing sub-assembly 104. In general, the circulatory system has several main purposes, including cooling and lubrication drill bit, landwaste is removed and is used mud from drill bit and pit shaft The wall of cake coating pit shaft.The circulatory system is made of drilling fluid, circulates through pit shaft downwards in drilling program.In general, following The component of loop system includes drilling fluid pump, compressor, associated conduit accessory and special to drilling fluid for adding additive Door syringe.In some implementations, such as, such as during horizontal or directional drilling program, downhole electrical motor is in combination with shaft bottom group Component 118 is used or is used in shaft bottom sub-assembly 118.Such a downhole electrical motor can be the mud motor with turbo arrangement Or the mud motor with the configuration of progression chamber, such as Moineau motor.These motors receive the drilling fluid across drill string 106 And it rotates to drive drill bit in drilling operation or change direction.

Many rotary drilling operation in, drilling fluid along drill string 106 downwards pump and pass through drill bit in mouth or Spout leaves.Then the annular space between pit shaft part 108 and drill string 106 (for example, annular space) is interior upwardly toward ground for fluid Face 102 is flowed, and the landwaste of suspension is carried to ground.Drilling fluid (pole is as drill bit) can be found according to below ground 102 The type of geological conditions selects.For example, the specific geological conditions and some subterranean layers that are found may need liquid (such as water) As drilling fluid.In these cases, it may be necessary to which water more than 100,000 gallons completes drilling operation.If water sheet Body is unsuitable for carrying out drill cuttings drilling or density is not enough to control the pressure in well, then can be (swollen by clay additive Profit soil) or water is added to form drilling fluid (for example, drilling mud) based on the additive of polymer.As described above, may There are problems that in relation to these additives may be used in the subsurface formations adjacent or close to the subterranean layer containing fresh water.

In some implementations, it drills sub-assembly 104 and shaft bottom sub-assembly 118 can be with the air or foam as drilling fluid Cooperation.For example, compressed air promotes the landwaste generated by drill bit vertically upward across annular space in air rotary drilling program To ground 102.Large-scale compressor can provide air, is pressed downward then along drill string 106 and eventually passes through the osculum in drill bit Or spout evolution.The landwaste removed to ground 102 is then collected.

As described above, the selection of drilling fluid can be according to the type of the geological stratification met with during drilling operation.In addition, this A decision may be influenced by probing type (such as vertical probing, horizontal drilling or directional drilling).In some cases, example Such as, specific geological stratification may be more suitable for air probing in vertical probing than orientation or horizontal drilling.

As illustrated in fig. 1, shaft bottom sub-assembly 118 (including drill bit) drills out or is formed vertical bore part 108, from ground Face 102 extends towards target subterranean zone 124 and producing formation 126.In some implementations, target subterranean zone 124 can be for suitable for sky The geological stratification that pneumatic drill is visited.In addition, in some implementations, producing formation 126 can be the geological stratification for being less suitable for air drilling program. As illustrated in fig. 1, producing formation 126 is close to formation at target locations 124 or below formation at target locations 124.Alternatively, in some implementations In, there may be subterranean layers among one or more (for example, different rock-layers between target subterranean zone 124 and producing formation 126 Or mineral layer).

In some implementations of slanted well bore system 100, vertical bore part 108 can be cased with one or more casings.Such as Diagram, vertical bore part 108 include guide sleeve 110, are extended in ground simplely from ground 102.By guide sleeve 110 A part of the vertical bore part 108 sealed can be major diameter pit shaft.For example, this part of vertical bore part 108 can For the 17-1/2 " pit shaft with 13-3/8 " guide sleeve 110.In addition, in some implementations, vertical bore part 108 can deviate Vertical plane (for example, slanted well bore).Further, in some implementations, vertical bore part 108 can be scalariform pit shaft, so that A part bores vertically downward and is then bent into substantially horizontal pit shaft part.Substantially horizontal pit shaft part can then turn downwards To the second generally vertical component, turning next to the second generally horizontal wellbore part.Additional generallyperpendicular and water Horizontal well canister portion point can be according to such as type on ground 102, the depth of one or more target subterranean zones, one or more grown places The depth of lower layer and/or the addition of other standards.

The underground direction of guide sleeve 110 can be surface pipe 112.Surface pipe 112 can seal slightly smaller pit shaft and Protect invasion of the vertical bore part 108 not by the fresh water aquifer for example near ground 102.Vertical bore part 108 Can then it extend vertically downward towards kickoff point (KOP) 120, it can be above target subterranean zone 124 between 500 feet and 1000 feet. This part of vertical bore part 108 can be sealed by intermediate casing 114.Vertical bore part 108 is any in its length The casing size of diameter and any of above casing at point can be the suitable dimension according to drilling program.

When reaching kickoff point (KOP) 120, boring tool (such as well logging and measuring device) can be deployed to pit shaft part 108 In.In this regard, the determination of the exact position of shaft bottom sub-assembly 118 can be made and be transmitted to ground 102.In addition, reaching When kickoff point (KOP) 120, shaft bottom sub-assembly 118 can be varied or adjusted, so that suitable directional drilling tool can be inserted into vertically In pit shaft part 108.

As illustrated in fig. 1, crooked hole canister portion points 128 and horizontal wellbore part 130 are in one or more geological stratifications It is formed.In general, crooked hole canister portion points 128 can be bored since the downhole end of vertical bore part 108, and from vertical bore part 108 towards preset bearing angular variation, and the 100 feet of gains of every brill are between 9 degree and 18 degree of angles.Alternatively, different preset bearing Angle can be used for boring crooked hole canister portion point 128.When boring crooked hole canister portion point 128, shaft bottom sub-assembly 118 is usually using measurement while drilling (" MWD ") equipment more accurately determines position of the drill bit in one or more geological stratifications (such as target subterranean zone 124).It is logical Often, MWD device can be used for orienting starter when drill bit forms crooked hole canister portion point 128 and horizontal wellbore part 130.

As the substituted or supplemented of MWD data is compiled during drilling out pit shaft part shown in Fig. 1, pit shaft can drilled out Specific high-fidelity measurement (for example, exploration) is carried out during part.For example, above can periodically be surveyed (for example, specific the time Drill the duration), periodically surveyed in pit shaft length (for example, drilling out specified distance, such as every 30 feet or It is other) or as needed or when the problem of (for example, when there are closing well cylinder path) is surveyed in expectation.In general, in the exploration phase Between, the gradient and azimuthal complete well measurements (in general, total length when measurement) of position in well are carried out according to rationally Accuracy understanding follows correct or specific pit shaft path (for example, planning according to pit shaft).In addition, the case where must bore relief well Under, it may be beneficial for understanding position.High-fidelity measurement may include pit shaft from the gradient of upright position and azimuth (or compass Course), on condition that path direction is crucial.It can carry out these high-fidelity measurements at the discrete point of pit shaft, and from discrete point Calculate the approximate path of pit shaft.High-fidelity measurement can be carried out with any appropriate high-fidelity sensor.Example includes for example simple Pendulum shape device to complicated electronic accelerometer and gyroscope.For example, in the measurement of simple pendulum, freely-suspended pendulum Position relative to examination network (be attached to the shell of measuring tool and being assumed represent pit shaft path) is trapped in film On.When tool is removed from pit shaft, on cable or when drilling pipe next time pulls out of hole from drilling, film is rinsed and checks.

Horizontal wellbore part 130 can usually extend up to hundreds of (if not thousands of) feet in target subterranean zone 124.Although figure 1 is illustrated as horizontal wellbore part 130 to be exactly perpendicularly to vertical bore part 108, it is to be appreciated that the pit shaft that orientation drills out is (all Such as horizontal wellbore part 130) there are some variations on its path.Therefore, horizontal wellbore part 130 may include " zigzag " road Diameter, but remain in target subterranean zone 124.In general, horizontal wellbore part 130 is drilled to predetermined end point 122, as described above Up to apart from 120 several thousand feet of kickoff point (KOP).As described above, in some implementations, crooked hole canister portion point 128 and horizontal well canister portion Points 130 can be using using air or foam to be formed as the air drilling program of drilling fluid.

Wellbore system 100 further includes the controller 132 communicated with BHA 118.Controller 132 can be located at well site at (for example, At probing sub-assembly 104 or near it) or can far connect well site.Controller 132 can also be with other systems, device, database And network communication.In general, controller 132 may include processor-based computer (for example, desktop PC, meter on knee Calculation machine, server, mobile device, mobile phone or other) comprising memory (for example, magnetic, light, RAM/ROM, can be removed, Remotely-or locally memory), network interface (for example, interface based on software/hardware) and one or more input/output periphery Equipment (for example, display device, keyboard, mouse, touch screen and other).

Controller 132 can at least partly control, manage and execute operation relevant to the drilling operation of BHA.In some sides In face, controller 132 can dynamic (for example, in real time) controls and adjusts pit shaft during drilling operation at wellbore system 100 The one or more of the illustrated component of system 100.Real-time control can be based on sensor measurement data or based on pit shaft track Change prediction, or even is adjusted without any sensor measurement.

Controller 132 can execute these control operations based on BHA dynamic model.BHA dynamic model analog drilling operation In various physical phenomenons, such as vibrational perturbation and sensor noise.It is pre- to determine that BHA dynamic model can be used in controller 132 It well logging cylinder track and adjusts one or more weighted factors and emphasizes or de-emphasize different target relevant to probing with selectivity.

In general, BHA dynamic model can rely on the base state variable of Temporal Evolution, represent in drilling operation Change condition.State variable in BHA dynamic model is the estimation of the time of day of BHA, from can wherein export pit shaft track Estimation.The dynamic temporal evolution of BHA can be indicated that the example may be formulated by discrete-time state-space model are as follows:

Wherein matrix A, B and C are the dynamic sytem matrixes in basis for indicating BHA probing and measurement.By in drilling program The basic physics and mechanism of middle use determine sytem matrix A, B and C.In practice, these matrixes are based on experience estimation and build Mould.State x (t) is vector, indicates the continuous state of BHA system, and input u (t) is the vector for indicating BHA control input, and Exporting y (t) is the vector for indicating observation (measurement) track of pit shaft.

In certain aspects, vector w (k) representation program noise and vector v (k) indicate measurement noise.Program noise w considers Such as factor of the effect of rock-drill bit interaction and vibration, and measure the noise in noise v consideration measurement sensor.It makes an uproar Sound path sequence w (k) and v (k) may be not accurately known, but can carry out Rational Conjecture to these programs, and these conjectures can be based on Experience modification.Noise vector w (k) and v (k) is usually by Gauss Procedure modeling, but non-Gaussian noise can also pass through modification state x With matrix A not only to include the dynamic by state-variable description, but also including random noise as described further below Dynamically model.

In the example being discussed below, it includes 6 control variables that BHA, which controls input vector u (t), indicates that the first of BHA is curved The starting of angle and the second bent angle, the depth of BHA, the first packer and the second packer is (for example, pass through the inflation of packer, envelope Every the mechanical compression etc. of device) and packer separation.Output vector y (t) includes 12 measured values observed, including from attached 6 measured values and encapsulate (hereinafter referred to as " inc/ from distal end inclinometer and magnetometer that nearly inclinometer and magnetometer encapsulate Mag ") it is other 6 measurement.State vector x (t) is dimension 12+n_dVector comprising represent true bearing angle and gradient 12 states of value such as will encapsulate observation (measurement) by attached proximal and distal end inc/mag.Value n_dIt is the magnitude of Disturbance Model, Filter unmodeled disturbance and total dynamic 12 states of expression system.

Therefore state transition matrix A is (12+n in this example_d)×(12+n_d) dimension state transition matrix, indicate base Plinth physics, matrix B are (12+n_dMatrix, the relationship between control control variable and system mode are tieed up in) × 6, and Matrix C is 12×(12+n_d) matrix, manage the relationship between observation y and system mode x.Matrix A, B and C, which can be used, any suitably to be estimated Meter or modeling technique (such as lumped mass system model) determine.If more complicated dynamic model is used to describe system, There may be more multimodes.

Random noise and potential inaccuracy when due to being modeled to sytem matrix A, B and C, the BHA dynamic in equation 1 The state x of model is not in general accurately known, but be inferred to.Under these scenes, equation 1 can be used for determining shape The inferred value or estimated value of state x and measurement y, rather than their true value.Particularly, the model of equation 1 can be used for generating The prediction of the future value of state x and observation y.The model dynamic that these predictions are contemplated that actual measurement to improve in equation 1.

For example, without any current measurement,Following equation can be used to obtain next shape of BHA system The estimated value of state:

If currently measurement y can be obtained, prediction can be generated by using Kalman filtering renewal equation formula:

In equation 3, y (k) indicates actual observation (for example, measuring by high-fidelity sensor measurement, MWD sensor Or any other proper sensors measurement provides).Factor K (for example, when variable factor) (also referred to as Kalman observes gain) table Show for considering actual path and estimating the correction factor of the error between track.In general, K's is larger Value is implied in the estimated value for determining NextState, and bigger weight is assigned to the observation y (k) measured.In general, Quantity of the K according to the vibration and reaction force for influencing drill bit.K value can be according to any proper standard (for example, minimizing state estimation The mean square error of value or any other proper standard) selection with realize measurement observation and basic model it is dynamically relatively heavy Expectation tradeoff between the property wanted.

BHA dynamic model in equation 1 can receive new information by controller (for example, controller 132 in Fig. 1) When dynamic update.For example, matrix A and B are influenced by the operating condition in pit shaft, because model (such as in equation 1) can be with behaviour Make condition to change and linearize again.For example, can be when determining that BHA enters different subterranean layers, or in drilling operation from straight line Probing direction linearizes again as executing when changing direction into bending drilling trajectory.In general, when drilling environment changes When, BHA dynamic model can be updated for various reasons.

The BHA dynamic model in equation 1 can be used based on the predictive controller of model to generate the following pit shaft track It predicts and predicts to determine that the BHA input control for meeting (such as optimization) expectation target function also meets one simultaneously based on these Or multiple constraints.Objective function can be the combination of at least one target weighted by weighted factor.It can be based on measurement, prediction Weighted factor is dynamically adjusted in response to changing the condition in pit shaft with other information.

As illustrative example, objective function can minimize the weighted array of two targets in future horizon: (1) and advising Draw the deviation in pit shaft path, and the input energy that (2) are consumed according to constraint set by BHA.One of such example goal function Example stratum is shown in following equation 4 and 5:

Foundation

Wherein y^spIt is planning pit shaft path, t indicates current time and T is that (it may be limited to obtain prediction time domain Dynamic resolution may be unlimited to obtain steady state solution).The first item in objective function in equation 4 is quadratic term, right Ying Yu makes the target minimized with the square deviation in planning pit shaft path, by that can weight for the weighting matrices Q (k) of time-varying. Section 2 in equation 5 is quadratic term, corresponds to the target for minimizing square variation with input control, indicates logical Cross the input energy consumption of weighting matrix S (k) (it can be time-varying) weighting.In Section 2, it is assumed that underground power consumption and input control Rate of change (for example, starting of bent angle and packer) it is proportional.The variation of input control is that control is inputted in consecutive hours spacer step Difference between system.Δ u (k)=u (k)-u (k-1) function G () is that input-output indicates, based on the BHA in equation 1 Dynamic model.Particularly, equation 2 (for the update in the case where no measurement) or equation 3 can be used in function G () (for example, for the update for having measurement) is predicted with the next step for generating measurement y based on desired BHA input control u.

In current time step t, the objective function in equation 4 is being solved to generate desired control signal sequence u (k), k =t, t+1 ..., after (t+T), only first control signal u (t) is applied to BHA.In subsequent time t+1, equation is solved again Objective function in formula 4 to generate next control sequence u (k), k=t+1 ..., (t+1+T), wherein controlling u (t+1) for first It is applied to BHA.These iteration continue, and eyes front T step should be applied to BHA to generate best current step control u with the side of satisfaction Objective function in formula 4.It in each iteration, can be based on the weighted factor in measurement and forecast updating matrix Q and S to adjust At the change condition in pit shaft.

In some instances, weighting matrices Q can be diagonal form, wherein different weights are distributed by item diagonally To 12 inclinometers and magnetometer measures.Gained control is made great efforts true by the machinery and formation properties of the lithosphere in drilling environment It is fixed.For example, the weight for assigning each measurand is based on its steady-state gain if Q is unit matrix.But one In a little examples, it may be desirable to adjust the weighted factor in Q with selectivity and emphasize that (or de-emphasizing) specifically measures.Particular measurement variable The instruction of larger weighted factor BHA input should be designed in a manner of forcing the particular measurement variable and more reinforce (more accurate) control Control.On the contrary, the smaller weighted factor instruction of particular measurement variable should be to allow more loosening (less more for the particular measurement variable Accurately) mode controlled designs BHA input control.

In general, objective function is without being limited to weighted factor being expressed as coefficient such as in the example of equation 4 Equation.More approximately, objective function can indicate any appropriately combined of one or more targets, and weighted factor can be with Indicate any appropriate quantization of the compromise between different target.For example, if objective function includes reducing and planning pit shaft path Deviation and reduce input energy target, then weighted factor can substantially indicate the compromise between deviation and energy.

Solving objective function can be related to the control for the expectation compromise that wherein weighted factor is used to select to realize between different target Make any proper technology of input, such as iterative technique, numerical technique or heuristic technique (or other proper technologies).For example, If objective function (belt restraining) is expressed as equation 4 and 5, solution can be related to any appropriate Optimization Solution technology.Make For another example, solving objective function can be related to the step of causing desired control to input series.For example, two step programs can include: First, it can get and realize minimum or control input group close to the candidate BHA of minimum input energy (for giving set of constraints)；With Two, it can be from the input of the expectation deviation in candidate input group selection realization and planning pit shaft path.It can be used relative to input energy The maximum of the appropriate compromise quantization (for example, ratio between energy and deviation), given minimum input energy that indicate weighted factor is partially Poor or some other compromise conceptual choices it is expected deviation.

The example that Fig. 2 illustrates the process flow based on Model Predictive Control of BHA, in response to the change condition in pit shaft And dynamic adjusts weighted factor.It can be for example by the controller of BHA (for example, BHA 118 in Fig. 1) (for example, the control in Fig. 1 Device 132) carry out Fig. 2 instance processes process 200.In the example of figure 2, in box 202, objective function is solved (for example, side Objective function in formula 4) to generate control input 204 to BHA206 (for example, BHA 118 in Fig. 1).Objective function can It with the model (for example, model in equation 1) based on BHA dynamic 208 and may include by one or more weighted factors 210 Any amount of appropriate target of (for example, weighted factor in matrix Q and S in equation 4) weighting.

In the example of figure 2, the sensor measurement 212 from BHA can be used to update box 202 via measurement feedback 214 In objective function solve.Furthermore, it is possible to update one or more of the other part of drilling operation based on sensor measurement 212. For example, can dynamically adjust weighting matrices Q using the measurement 212 from BHA based on the change condition in pit shaft in box 216 With the weighted factor in S.Further, it is also possible to update the dynamic model of BHA based on sensor measurement 212 in box 218.These Dynamic, which updates, can support BHA input control 204 to adapt to the complicated variation in subsurface environment.Make a reservation for during the design phase with using Constant-weight matrix compare, these adjustments can support more accurate BHA control inputs and more effectively whole probing Operation.

In some instances, weight adjustment box 216 can be used sensor measurement 212 and determine the uncertain of pit shaft track Property, and may then based on identified uncertain adjustment weighted factor (for example, the weighting of matrix Q and S in equation 4 The factor).In some instances, weight adjustment box 216 can be additional or uses BHA dynamically and state renewal equation formula The model of (for example, state renewal equation formula 2 and/or 3) determines following probabilistic prediction.Weight adjusts box 216 can be with Uncertain adjustment weighted factor based on the pit shaft track in specific direction is that more will emphasize or less emphasize that specific BHA is defeated Enter control.

As example, the vibration in drilling program may be due to the operation probing in the different piece of pit shaft and in not Tongfang To middle generation.This vibration can increase the uncertainty of pit shaft track.It can wish to increase when uncertain dynamic increases special Determine weighted factor in direction (or dynamic reduces weighted factor when uncertainty reduces).Weight adjustment box 216 can be based on Measurement automatically determines uncertainty based on model prediction, and correspondingly adjusts weighted factor.

In addition, in the example of figure 2, weight adjusts the mesh in box 216 and model modification box 218 and box 202 Scalar functions solver is also adapted for other information, such as pit shaft planning information 220.As example, planning information 220 may include Plan pit shaft path and the information about other pit shafts in the near wellbore.In some instances, weight adjusts box 216 And/or model modification box 218 can also be using planning information 220 to update weight (for example, matrix Q and S in equation 4 In weighted factor) and BHA dynamic (the dynamic model of BHA in equation 1) model.

Weight, which adjusts box 216, synthesizing for weight adjustment mechanism to update adding in matrix Q and S based on one or more Weight factor, the weight adjustment mechanism is based on measurement described above, prediction and/or planning information.These weighted factors update Can be occurred with any appropriate time scale, such as every time stamp or optionally every time measurement when.

Be provided below weight adjustment mechanism some examples, but also can be used with determine weighted factor (and therefore BHA control System input) relevant other adjustment mechanism.For example, can be based on the torque or base on the measurement feature such as BHA of down hole drill Weight matrix Q and/or S are adjusted in such as Angle Position of fluid flowing or downhole tool of the constraint in BHA input.In general, Any for drilling operation is properly entered or exports, and can define one or more weighted factors with appropriate adjusting and be placed on this Constraint in input or output.

Examples below describes three feasible adjustment mechanism using three different types of information dynamically to adjust weighted factor. These adjustment mechanism are to be based on: uncertainty, planning pit shaft path and the anti-collision hitting information of pit shaft track.

As the first example of weight adjustment mechanism, weighted factor can be determined based on uncertainty.In general, sensing Device measurement (for example, MWD data, survey data etc.) can improve the accuracy of pit shaft tracking and track assessment.However, practical On, presence (for example, reaction force from rock stratum, vibration etc.) the creation sensor measurement of sensor noise and program noise It is uncertain.The uncertainty of the pit shaft track of measurement can be with covariance matrix Σ_yIt is characterized.In some instances, matrix Σ_yIt can be by calculating the covariance between the multiple measurement orientation values for the pit shaft track collected with the time and measurement tilting value Value determines.It in some instances, can be by using pit shaft rail other than using sensor measurement or alternatively The prediction of mark determines matrix Σ_y(determining for example, such as being updated by the BHA dynamical state in equation 2 and/or 3) is to generate not Carry out probabilistic prediction in pit shaft track.

In this example, covariance matrix Σ_yDiagonal entry be side in each of 12 inc/mag measurement Difference.Covariance matrix Σ_yThe outer element of diagonal line be that different inc/mag measure covariance value between, described in description Correlative between measurement.

In the above example, if sensors measure that y have in particular directions it is a large amount of uncertain (for example, due to Rise in the big vibration force in the probing direction), then this, which is indicated generally at, measures more unreliable and true pit shaft track in the direction It is upper that there is big error tolerance.In this scene, it may be desirable to be exerted on probabilistic direction using more multiinput control Power.Furthermore it or substitutes, it may be desirable to which less emphasis is maintained upward by between measurement track and planning pit shaft path in uncertainty side Little deviation (since the track of measurement is more unreliable in this direction).In terms of weighted factor, this can be by reducing adding for output Weight factor (for example, weight matrix Q in the first quadratic term of equation 4) or the weighted factor by increasing input are (for example, side The weight matrix S in second secondary item in formula 4) and implement.In some instances, the weighted factor for increasing input can correspond to In the constraint (for example, reducing peak swing constraint, reducing mean power constraint etc.) for reinforcing inputting.Similarly, in some realities In example, the weighted factor that reduces output, which can correspond to, loosens the constraint in output (for example, increasing and the deviation of planning pit shaft track The probability etc. of maximum boundary, increase and planning pit shaft trajector deviation beyond predetermined amount).

For example, the adjustment of quantization output weighted factor Q is to explain a kind of probabilistic method:

Q=Q₀∑_y ^-1 (6)

Wherein Q₀It is the relative Link Importance of each output (measurement).In some instances, Q₀It can be set to unit matrix I₁₂, but Q₀It can be any appropriate matrix for assigning weighted factor to different measurement directions.The Matrix Formula of equation 6 is supported Along any cross-wise direction application input control, without being limited to any specific direction axis.

Fig. 3 illustrates the probabilistic 3 dimension example of the correlation between the direction in pit shaft track.In this example, horizontal Pit shaft 300 has the track along the probing direction 302 for being parallel to the first axis direction 304.The second axis is determined based on sensor measurement Direction 306 and third axis direction 308 are both influenced by probing vibration.As example, exporting uncertainty can be expressed as assisting Variance matrixIn this example, diagonal line value (1 and 6) is the covariance of the second axis and third axis.Diagonally Value (2 and 2) is the covariance (related) between the second axis and third axis outside line.Due between the second axis and third axis it is related with Correlation between third axis and the second axis is identical, so the outer element of diagonal line is identical.

Fig. 3 shows this explanation, wherein ellipse 310 indicates one between the second axis direction 306 and third axis direction 308 It causes uncertain and related between 312 the second axis direction 306 of expression of orientation correlation and third axis direction 308.Pass through equation 6, use initial weight matrix Q₀Corresponding output weight matrix is determined as by=IThe outer element of diagonal line is built Discussing MPC not only should control BHA along the second axis direction 306 and third axis direction 308 during probing, it is also contemplated that the second axis direction The friendship indicated between third axis direction (alternatively, in general, between any two main shaft) by the orientation related 312 in Fig. 3 Fork is related.This example illustrates how the weight adjustment calculated based on covariance matrix is supported to input BHA in any direction It is uncertain that control is suitable for track.Such adjustment can reduce the correlation in uncertainty, and help prevent along a side Vibration to the other directions for influencing probing.

Other than adjustment exports weight matrix Q or alternatively, input weight matrix S may be adapted in uncertainty Variation.It is mobile that adjustment input weight matrix S can inhibit or amplify BHA input control.For example, not known when pit shaft track Property it is larger in particular directions when, it may be desirable to use more BHA control to make great efforts (to improve in the direction in the specified direction The accuracy of upper control BHA), and weight associated with specific input direction can be adapted these variations of automatic implementation. As example, following equation may be used to determine the mobile weight matrix S of input:

WhereinIndicate model steady-state gain (for example, steady-state value of the input and output gain G () in equation 5) and OperationIndicate pseudoinverse (since the transfer function matrix of drilling program can be non-square, that is, control the amount and output of variable Measurement can be unequal).Equation 7 is translated into the contribution of overall output by pit shaft uncertainty is exported according to each input variable BHA input control effort.As illustrative example, it is assumed that inclination sensor is by two bent angle controls near in third axis direction System, and assume steady-state gainPseudoinverseMean the inclination measurement in third axis direction Unit uncertainty causes the first bent angle and the second bent angle to have the mobile weight 0.2 and 0.4 of input respectively.This result and intuitive one It causes, because the influence of the second bent angle is twice of the influence of the first bent angle as proposed by steady-state gain.

Other than using the uncertainty of pit shaft track or alternatively, what is be considered as when determining weighted factor is another Factor is the shape of pit shaft.In some instances, pit shaft path is provided in design phase determination or by higher level's decision making algorithm. In such scene, planning pit shaft path may be used as feed-forward information to adjust output weight matrix Q.When planning pit shaft When path has zig zag, that is, radius of curvature becomes smaller, Random Effect, and such as lateral force can become stronger, causes more very much not Certainty matrix.In some instances, instead of waiting drill bit to become to deviate and start due to side force rotation (and therefore not really It is qualitative to become much larger), feed forward mechanism can be used for shifting to an earlier date actively adjustment output weight matrix Q based on planning pit shaft routing information. Feed forward mechanism can be based on model or can be data-driven.The example of feedforward arithmetic based on model is by output weight Matrix indicates:

Q=diag { R₁,R₂, R₃} (8)

Wherein R₁、R₂And R₃Indicate the curvature half for being respectively relative to the first axis direction, the second axis direction and third axis direction Diameter.In some instances, weight matrix Q can correspond only to one group of inc/mag measurement.All 12 inc/mag are measured, Four in Q matrix can be stacked on to diagonal line in single bigger matrix.Alternatively, the feedforward arithmetic in equation 8 can be by Data-driven.For example, the historical sensor measurement data of the past part (for example, several hundred feet) of probing can be used to curvature Relationship between radius and uncertain matrix is modeled.

It is anti-collision safety information that weight, which adjusts another factor that algorithm considers,.When pit shaft is close to other existing pit shafts When, there may be the risk for piercing other pit shafts and causing collision.In such scene, it is usually desirable to be sent out along wherein most probable The direction of raw collision utilizes the control more reinforced.

The example that Fig. 4 A and Fig. 4 B illustrate the anticollision direction of determining weighted factor adjustment.In Figure 4 A, pit shaft 400 has There is the track along the probing direction 402 for corresponding to the first axis direction.Exist in pit shaft 404 relative to by 406 institute of fixed-course detector The direction of the pit shaft 400 of instruction.In this example, collision elimination direction 406 is not exclusively along one in the main shaft.This It illustrates in figure 4b, the example for showing the two-dimensional cross sectional figure of anticollision scene.In figure 4b, by directionality indicator 412 Collision elimination direction between the pit shaft 408 and 410 of instruction does not exist along the second axis direction 414 or third axis direction 416.At this In the scene of sample, weight matrix Q can have element outside the diagonal line for non-zero.

Weight matrix can be designed in any appropriate manner to reflect collision elimination information.As illustrative example, in Fig. 4 B In two-dimensional example can have the weight matrix Q being expressed as follows:

Wherein W is the weight relative to other (orthogonal) directions along collision elimination direction 406.θ in the example of equation 9 Angle, θ is collision elimination angle (for example, in polar coordinates).Example in Fig. 4 B is simplified example, and example for illustrative purposes If as, there are another pit shaft, General Principle can extend to three-dimensional space in 402 front of probing direction.

Weight synthesizer can be used to combine the above three factor.Particularly, the weighting square determined in equation 6,8 and 9 Battle array Q can synthesize the single output weight matrix Q that can be applied to objective function (such as objective function in equation 4).Appoint What appropriate synthetic technology can be used to such as by using weighted array and in composite equation formula 6,8 and 9 different matrixes (and/ Or the other matrixes determined by other adaptation techniques).Similarly, in equation 7 determine matrix S can with by other adjustment skills Other weight matrixs combination that art determines.Other adaptation techniques may depend on any appropriate measurement relevant to drilling operation, pre- Survey or planning information.

Fig. 5 illustrates the flow chart (such as weight adaptation module 216 in Fig. 2) of the program of weight adjustment and synthesis.In In the example of Fig. 5, weight adjusts program 500 and uses both feedforward function 502 and feedback function 504.Feedforward function 502 can make With any appropriate pit shaft planning information (for example, planning information 220 in Fig. 2), the example includes planning pit shaft path 506 and prevents Collision information 508.Any appropriate feedback information 510 generated due to drilling operation can be used in feedback function 504.

Feedback information 510 may include the uncertainty of the pit shaft track for example such as generated by uncertainty models 512.No Deterministic models 512 can determine the uncertainty of pit shaft track based on drilling information 514, may include that measurement and/or control become Amount (for example, sensor measurement 212 and/or control variable 204 in Fig. 2).For example, uncertainty models 512 can pass through calculating Covariance value between sensor measurement generates uncertain prediction by using BHA dynamic model and determines covariance matrix (for example, the Σ in equation 6 and 7_yCovariance matrix).

Feedback function 504 is measured using uncertainty and/or is predicted to generate feedback weight information 516.It is grasped by synthesis Make 520 to be synthesized together feedback weight information 516 and feedforward value information 518 by feedforward function 502, to generate weighted factor 522 (for example, weight matrix Q and S in equation 4).In some instances, weight synthesis program 500 is in future prediction horizon Upper output weighted factor 522 (for example, infinite horizon of the finite time-domain T in equation 4 or steady state solution).Weighting coefficient 522 is right It is input into the PREDICTIVE CONTROL 524 (for example, such as being implemented by the objective function solver in the box of Fig. 2 202) based on model afterwards, To generate BHA input control.

After obtaining the input/output data drilling information 514 from drilling operation, more by uncertainty models 512 The uncertainty of new pit shaft track, the uncertainty are sent back to feedback function 504 to update to feedback weight information 516. Meanwhile planning that pit shaft path can be updated and be input to feedforward function 502, before the planning information based on update recalculates Present value information 518.

Fig. 6 is the flow chart for the example procedure 600 for executing the PREDICTIVE CONTROL based on model of BHA.The example journey of Fig. 6 One or more steps of sequence can be executed by pit shaft controller (for example, controller 132 in Fig. 1).In this example, it controls Device processed determines the sensor measurement (602) from BHA.Controller determines BHA dynamic model based on the sensor measurement from BHA (for example, model in equation 1) (604).Then controller determines the weighted factor for corresponding to Target For Drilling (for example, equation The weighted factor in matrix Q and S in formula 4) (606).Controller determination includes the Target For Drilling and one weighted by weighted factor The objective function (for example, objective function in equation 4) (608) of a or multiple constraints (such as constraint in equation 5).Control Then device processed determines meets the objective function (for example, objective function in optimization method formula 4) and one or more constrains BHA control input (610), and control input is applied to BHA (612).

Fig. 7 be determine correspond to Target For Drilling weighted factor (for example, weighting in matrix Q and S in equation 4 because Son) (for example, the flow chart of the example of the further details of the step 606) in Fig. 6.In this example, controller determine with At least one of lower item: the uncertainty of pit shaft track, the shape of pit shaft or the collision elimination information (700) of measurement.So Afterwards, at least one of the uncertainty of pit shaft track, the shape of pit shaft or the collision elimination information of controller based on measurement is true Determine weight (for example, weight of weighting matrices Q and/or S in equation 6 to 9) (702).Then controller closes the weight As weighted factor (for example, the weighting matrix determined in equation 6,8 and 9, which can synthesize, can be applied in equation 4 The single output weight matrix Q of objective function) (704).

Fig. 8 be determination include by weighted factor weighting Target For Drilling and one or more constraint objective function (for example, The flow chart of the example of the further details of step 608) in Fig. 6.In this example, controller, which determines, carrys out self planning pit shaft The following deviation (800) of the prediction in path.Controller also determines the future cost that control input is applied to the prediction of BHA (802).Then controller determines that the following deviation for the prediction for coming self planning pit shaft path is applied to the pre- of BHA with that will control input The weighted array (804) of the future cost of survey weighted by weighted factor.

Fig. 9 be determining objective function (for example, objective function in equation 4) and determine BHA control input into one Walk the flow chart of the example of details, the objective function include by weighted factor weighting Target For Drilling and it is one or more about Beam (such as the step 608) in Fig. 6, the control input meet the objective function and one or more of constraints (such as Step 610) in Fig. 6.In this example, controller determines the following deviation for the prediction for coming self planning pit shaft path and will control System input is applied to the weighted array of the future cost of the prediction of BHA (for example, the step 804) in Fig. 8.Controller then with Determined on period time afterwards minimize (such as equation 4 in) come self planning pit shaft path prediction the following deviation and will Control input is applied to the BHA control input (900) of the weighted array of the future cost of the prediction of BHA.

Figure 10 is the block diagram of the example of computer system 1000.For example, with reference to Fig. 1, the one or more of controller 132 The example that part can be system 1000 as described herein is such as used by any user of the resource of access wellbore system 100 Computer system.System 1000 includes processor 1010, memory 1020, storage device 1030 and input/output device 1040.Component 1010,1020,1030 and 1040 can be interconnected for example using system bus 1050.Processor 1010 is capable of handling Instruction in system 1000 for executing.In some implementations, processor 1010 is single-threaded processor.In some implementations, Processor 1010 is multiline procedure processor.In some implementations, processor 1010 is quantum computer.Processor 1010 can be located Reason is stored in memory 1020 or is stored in the instruction on storage device 1030.Processor 1010 can execute operation, such as It determines the dynamic model of BHA, determine weighted factor, the control input for determining BHA etc. (such as Fig. 6 to 9).

Memory 1020 stores information in system 1000.In some implementations, memory 1020 is computer-readable Medium.In some implementations, memory 1020 is volatile memory-elements.In some implementations, memory 1020 is non-easy The property lost memory cell.

Storage device 1030 can provide massive store for system 1000.In some implementations, storage device 1030 is Computer-readable medium.In various different implementations, storage device 1030 may include such as hard disk device, optical disc apparatus, solid-state Driver, flash disk, tape or some other mass storage devices.In some implementations, storage device 1030 can be deposited for cloud Storage device, it may for example comprise be distributed in the logical storage devices of multiple physical storage devices on network and using network access. In some instances, storage device can store long term data, such as rock stratum data or ROP designed capacity.Input/output device 1040 provide input/output operations for system 1000.In some implementations, input/output device 1040 may include one or more A Network Interface Unit, for example, Ethernet card, serial communication apparatus (such as the port RS-232) and/or radio interface device, example Such as 802.11 cards, 3G radio modem, 4G radio modem or homing pigeon interface.Network Interface Unit allows system 1000 controllers 132 commuted in Fig. 1 communicate, such as transmission and reception instruction.In some implementations, input/output device May include driving device, the driving device be configured to receive input data and send output data to it is other input/it is defeated Device out, such as keyboard, printer and display device 1060.In some implementations, mobile computing device, mobile communication can be used Device and other devices.

Server (for example, the server for forming a part of controller 132 shown in Fig. 1 or wellbore system 100) can It is realized by instruction, described instruction causes one or more processing units to execute above procedure and function when being executed, for example, all The control input for meeting objective function is such as generated, relation information is generated, relation information is sent to BHA and application by relation information Control input of instruction etc. is (for example, Fig. 6 to Fig. 9).These instructions may include for example interpreted instruction, such as directive script or can Execute the other instructions of code or storage in computer-readable medium.The different components of wellbore system 100 can divide in network Cloth is implemented (such as server zone or one group of widely distributed server) or be may be implemented in multiple including what is intemperated with one another In the single virtual device of distributed devices.For example, a device can control other devices or device can be in one group of co-ordination principle Or operation or device can be coordinated in another way under agreement.The coordinated manipulation of multiple distributed devices, which is presented, is used as single device The performance of operation.

Described feature may be implemented in Fundamental Digital Circuit or computer hardware, firmware, software or their combination in. Equipment may be implemented in computer program product, be tangibly embodied in information carrier (for example, machine-readable storage device In) for being executed for programmable processor；And method and step can be executed by programmable processor, the processor executes instruction journey Sequence is to execute the described function of implementing by operation input data and generation output.Described feature can be advantageously carried out In one or more computer programs, the computer program be can be executed on a programmable system, the programmable system packet At least one programmable processor is included, the programmable processor is coupled to from data-storage system, at least one input dress Set at least one output device receive data and instruction and transmission data and instruction to data-storage system, at least one is defeated Enter device and at least one output device.Computer program is one group of instruction, and described instruction can directly or indirectly be used in computer In to execute specific activities or lead to particular result.Computer program can be write with any type of programming language, including compiling Or interpretive language, and it can be disposed in any form, including or as stand-alone program or as module, component, subprogram be applicable in Other units in calculating environment.

Appropriate processor for executing instruction program for example includes general and special microprocessor and any type Computer separate processor or one of multiple processors.In general, processor will be stored from read-only memory or random access Device or both receives instruction and data.The element of computer may include processor for executing instruction and for storing instruction and One or more memories of data.In general, computer may also comprise or be operatively coupled to with file for storing data One or more mass storage devices communication；These devices include magnetic disk, such as internal hard drive and moveable magnetic disc；Magneto-optic Disk；And CD.Storage device suitable for tangible embodiment computer program instructions and data includes that the non-volatile of form of ownership is deposited Reservoir for example includes semiconductor memory system, such as EPROM, EEPROM and flash memory device；Disk is such as internal hard Disk and moveable magnetic disc；Magneto-optic disk；With CD-ROM and DVD-ROM disk.Processor and memory can be by ASIC (special integrated electricity Road) it supplements or is incorporated in.

It is interacted to provide with user, feature is implementable on computers, include for displaying information to user's Display device, such as CRT (cathode-ray tube) or LCD (liquid crystal display) monitor；And indicator device, such as mouse or track Ball, user can be provided by being input to computer.

Feature is implementable in computer systems comprising aft-end assembly, such as data server or it include middleware Component, such as apps server or Internet Server or it include front end assemblies, such as client computer, tool There are graphic user interface or Internet-browser or any combination thereof.System component can pass through any form or the digital number of medium It is connected according to communication (such as communication network).The example of communication network include for example, LAN, WAN and formed internet computer and Network.

Computer system may include client and server.Client and server is typically remote to be passed through each other and usually Network (all networks as described) interaction.The relationship of client and server is by means of running on the respective computers and having The computer program of mutual client-server relation generates.

In addition, the logic flow described in figure be not necessarily to shown in particular order or successively sequentially realize expected result.In addition, Other steps can be provided, or can from described process removal process, and can be added to described system or from its Remove other components.Therefore, other embodiment is in the range of following claims.

Several implementations have been described.However, it will be appreciated that can carry out various modifications.For example, the additional aspect of program 600 may include Than more step illustrated in Fig. 6 and Fig. 9 or less step.In addition, Fig. 6 the step of being illustrated into Fig. 9 can by in figure Shown in different order execute.Although concept can also be applied to other in addition, describing concept in pit shaft drilling system Program.For example, the other application for being inserted into conjunction with medical endoscope inspection or in which instrument and being controlled in circumstances not known.Cause This, other embodiment is in the range of following claims.

Claims

1. the computer implemented method that a kind of control shaft bottom sub-assembly follows planning pit shaft path, which comprises

Determine the sensor measurement from the shaft bottom sub-assembly；

Determine the weighted factor for corresponding to Target For Drilling；

Determination includes the objective function of the Target For Drilling weighted by the weighted factor and one or more constraints, wherein institute The system in future state that objective function includes the shaft bottom sub-assembly is stated, determines that the objective function includes the determining and planning The prediction future deviation in pit shaft path；

Determine that the control to the shaft bottom sub-assembly for meeting the objective function and one or more of constraints inputs；With

Control input is applied to the shaft bottom sub-assembly.

2. computer implemented method according to claim 1, wherein determining that the weighted factor for corresponding to Target For Drilling also wraps It includes:

At least one in the sensor measurement based on the shaft bottom sub-assembly dynamic model or from the shaft bottom sub-assembly A determining weighted factor.

3. computer implemented method according to claim 2, wherein based on the shaft bottom sub-assembly dynamic model or coming from At least one of described sensor measurement of the shaft bottom sub-assembly determines that weighted factor includes:

Determine at least one of shape or the collision elimination information of uncertain, the described pit shaft of the pit shaft track of measurement；

In the shape or the collision elimination information of uncertain, the described pit shaft of pit shaft track based on measurement At least one determines weight；With

The weight is synthesized into the weighted factor.

4. computer implemented method according to claim 3, wherein determining that the uncertainty of the pit shaft track of measurement includes Determine the covariance value between the multiple orientation values and tilting value of the pit shaft track.

5. computer implemented method according to claim 4, wherein determining that the weighted factor for corresponding to Target For Drilling includes Described in the pit shaft track of the measurement uncertain is reinforced to institute on increased direction wherein from previous measurement time State the constraint in the control input of shaft bottom sub-assembly.

6. computer implemented method according to claim 5, wherein reinforcing defeated to the control of the shaft bottom sub-assembly Constraint on entering includes determining the value added that associated weighted factor is inputted with the control to the shaft bottom sub-assembly.

7. computer implemented method according to claim 4, wherein the Target For Drilling includes and the planning pit shaft road The prediction deviation of diameter, and determine the weighted factor for corresponding to Target For Drilling include described in the pit shaft track of the measurement not really It is qualitative from previous measurement time wherein loosen on increased direction on the prediction deviation in the planning pit shaft path Constraint.

8. computer implemented method according to claim 7, wherein loosening the prediction with the planning pit shaft path Constraint in deviation includes the reduction of weighted factor determining and associated with the planning prediction deviation in pit shaft path Value.

9. computer implemented method according to claim 3, wherein determining that the shape of the pit shaft includes determining the rule Draw the radius of curvature of the further part in pit shaft path.

10. computer implemented method according to claim 9, wherein the Target For Drilling includes and the planning pit shaft road The prediction deviation of diameter, and determine that weighted factor includes the radius of curvature in the further part in the planning pit shaft path Reduce on the wherein direction of reduction from previous measurement time and the pact on the prediction deviation in the planning pit shaft path Beam.

11. computer implemented method according to claim 3, wherein determining that collision elimination information includes determining and another well The collision most probable of cylinder is in the direction wherein occurred.

12. computer implemented method according to claim 11, wherein the Target For Drilling includes and the planning pit shaft The prediction deviation in path, and determine that weighted factor includes in the direction that the collision most probable with another pit shaft occurs wherein Upper reinforcement and the constraint on the prediction deviation in the planning pit shaft path.

13. computer implemented method according to claim 4, wherein determining multiple orientation values of the pit shaft track and inclining Covariance value between inclined value further include:

Determine the received multiple azimuthal measuremenies of sensor from the shaft bottom sub-assembly and inclination measurement；With

Determine the received the multiple azimuthal measurement of the sensor from the shaft bottom sub-assembly and the association between inclination measurement Variance yields.

14. computer implemented method according to claim 4, wherein determining multiple orientation values of the pit shaft track and inclining Covariance value between inclined value further include:

Multiple bearing predictions and tilt prediction are determined based on the shaft bottom sub-assembly dynamic model；With

The covariance between the multiple bearing prediction and the tilt prediction is determined based on the shaft bottom sub-assembly dynamic model Value.

15. computer implemented method according to claim 1, wherein determining that objective function includes:

Determine the prediction future cost that the control input is applied to the shaft bottom sub-assembly；With

It determines and is weighted by the weighted factor, and the prediction future deviation for planning pit shaft path and by the control System input is applied to the weighted array of the prediction future cost of the shaft bottom sub-assembly.

16. computer implemented method according to claim 15, wherein determining that weighted factor includes:

First weighted factor of the prediction future deviation in determining and described planning pit shaft path；With

Determine the second weighted factor that the control input is applied to the prediction future cost of the shaft bottom sub-assembly.

17. computer implemented method according to claim 15, wherein determining the extremely well for meeting the objective function The control input of bottom sub-assembly includes determining the prediction minimized on the follow-up time period with the planning pit shaft path The following deviation and by it is described control input be applied to the shaft bottom sub-assembly the prediction future cost the weighted array To the shaft bottom sub-assembly control input.

18. computer implemented method according to claim 15, wherein control input is applied to the shaft bottom group The prediction future cost of component includes the prediction energy consumption of the shaft bottom sub-assembly.

19. computer implemented method according to claim 1, further include:

It determines to the candidate control input of the shaft bottom sub-assembly；

Pre- well logging is determined with the shaft bottom sub-assembly dynamic model based on the candidate control input to the shaft bottom sub-assembly Cylinder track；With

Pit shaft track and the deviation determination planned between pit shaft path and the planning pit shaft path based on the prediction Prediction future deviation.

20. computer implemented method according to claim 1, wherein determining that the control to the shaft bottom sub-assembly inputs packet It includes and determines at least one of the control of the first bent angle, the control of the second bent angle, the control of the first packer or the control of the second packer.

21. computer implemented method according to claim 1, further include:

Determine the sensor measurement of the update from the shaft bottom sub-assembly；

The shaft bottom sub-assembly dynamic model updated is determined based on the sensor measurement of the update from the shaft bottom sub-assembly；

Shaft bottom sub-assembly dynamic model based on the update or the sensor of the update from the shaft bottom sub-assembly are surveyed At least one of amount determines the objective function of the weighted factor and update that update；With

Weighted factor based on the update automatically adjusts the extremely shaft bottom sub-assembly for the objective function for meeting the update The control input.

22. computer implemented method according to claim 13, wherein determining the multiple azimuthal measurement and inclination measurement Between covariance value further include determine the uncertainty value in two different directions from the pit shaft track between Cross-correlation.

23. a kind of system that control shaft bottom sub-assembly follows planning pit shaft path comprising:

First assembly, rest on the ground or Near Ground；

Shaft bottom sub-assembly is at least partly placed in the pit shaft at subterranean zone or near subterranean zone, the shaft bottom combination Part is associated at least one sensor；With

Controller, is communicatively coupled to the first assembly and the shaft bottom sub-assembly, and the controller can be operated to carry out Including following operation:

Determine the sensor measurement from the shaft bottom sub-assembly；

Determine the weighted factor for corresponding to Target For Drilling；

Control input is applied to the shaft bottom sub-assembly.

24. a kind of non-transitory computer-readable storage media to include at least one computer program code of instruction, institute Instruction is stated to be operated when executed to promote at least one processor to carry out following planning pit shaft road for controlling shaft bottom sub-assembly The operation of diameter, the operation include:

Determine the sensor measurement from the shaft bottom sub-assembly；

Determine the weighted factor for corresponding to Target For Drilling；

Control input is applied to the shaft bottom sub-assembly.